应用线性光谱混合模型快速评价土地退化的方法研究

Application of Line Spectral Mixture Model to Rapid Assessment of Land Degradation in Semiarid Area

利用不同时相的ＴＭ 数据，采用线性光谱混合模型的方法，对内蒙古中西部半干旱地区的土地退化进行了以快速评价为目标的研究．在选用沙地、裸地、农耕地、牧草地作为基本组分的基础上，提取了反映土地退化的定量指标．相比其他遥感定量提取的方法，在干旱、半干旱地区进行土地快速分析和评价上。

We apply the Line Spectral Mixture Model (LSMM) to evaluate the land degradation in the western Inner Mongulia of China.Two temporal TM multispectral images of year 1987 and 1996 have been used.Four endmembers including desert、salina、agriculture vegtation and pasture land are defined to estimate the proportion of surface cover types(endmemders)within the pixel.Then we exstract the quantitative value of the degraded land by this method.Comparing to the other method,we discover that LSMM is a good method to extimat quickly land degradation in semiarid area.